🚀New paper by Meta
The paper presents "quality-tuning," a technique to improve the aesthetic quality of images generated by pre-trained text-to-image models by fine-tuning them with a small set of appealing images, achieving significant quality gains and outperforming existing models on benchmarks.
The paper presents "quality-tuning," a technique to improve the aesthetic quality of images generated by pre-trained text-to-image models by fine-tuning them with a small set of appealing images, achieving significant quality gains and outperforming existing models on benchmarks.
October 2, 2023 at 9:14 PM
🚀New paper by Meta
The paper presents "quality-tuning," a technique to improve the aesthetic quality of images generated by pre-trained text-to-image models by fine-tuning them with a small set of appealing images, achieving significant quality gains and outperforming existing models on benchmarks.
The paper presents "quality-tuning," a technique to improve the aesthetic quality of images generated by pre-trained text-to-image models by fine-tuning them with a small set of appealing images, achieving significant quality gains and outperforming existing models on benchmarks.